Diversity of Vegetation Dominated by Selected Grass Species on CoalMine Spoil Heaps in Terms of Reclamation of Post-Industrial Areas

S Grasses have a considerable potential for the adaptation to various, often extreme, habitat conditions. The aim of the work was to present the vegetation diversity of the coal-mine spoil heaps with the dominant share of grasses and to identify the main factors responsible for this diversity in the aspect of post-industrial land reclamation. The communities differ in reference to the species preferences to light, moisture, soil fertility and reaction, which is reflected in the wide variety of microhabitats in the area. It was shown that the increase in the abundance of certain grass species, including Calamagrostis epigejos, Festuca rubra, Festuca arundinacea, Phragmites australis, has a significant negative impact on the species richness, species diversity and the uniformity of distribution of species of the plant community. Preliminary analyses revealed that on post-mining waste, the biomass production of the dominant species is negatively correlated with biodiversity. The knowledge about the biology and ecology of grass species, as well as on the assembly rules may be used in the reclamation of degraded areas. Gaining the knowledge about the vegetation diversity of the coal-mine spoil heaps with the dominant share of grasses can be useful in planning the reclamation works, taking into account natural processes, which leads to the creation of a permanent vegetation cover at a given site, protecting it against water or wind erosion. In the future these areas may provide a number of important ecosystem services.


INTRODUCTION
Coal mine spoil heaps that were formed as a result of long lasting mining activity are an inseparable element of the Upper Silesian landscape.These sites are very specific habitats, often without soil and vegetation cover that is linked to the absence of a soil seed bank and lack or low amount of nutrients in the substratum.In addition, they can be characterized by high temperature, salinity, significant fluctuations in moisture, and thermal activity.Instability of the coarse grain substrate causes its susceptibility to the water or mass erosion [Maciak, 1999;Bradshaw, 2000;Woźniak, 2010;Chmura et al., 2011;Chmura et al., 2013;Molenda et al., 2013;Markowicz et al. 2015].All these factors contribute to the difficulty of colonizing them both by plant and animal species.However, the results of many studies revealed [Rostański and Woźniak, 2007;Woźniak, 2010;Błońska et al., 2013;Nicia et al., 2014], that there are species which colonize this area during spontaneous succession and form plant communities and are frequently different from those recorded on other anthropogenic sites.The vegetation patches that occur on coal mine spoil heaps are frequently dominated by single species that are characterized by high frequency and abundance, giving them a specific physiognomy.The dominant species have a set of functional traits [e.g.height, high seed production, parameters of leaves, biomass, root traits] that help them successfully compete with accompanying species and reduce their abundance in patches [Frieswyk et al. 2007; Ryś et al.,2016].Therefore, they play a crucial role in shaping both the abiotic [soil formation, nutrient and water cycling] and biotic components of plant communities [Simberloff and Von Holle, 1999].In reference to such species, a "mass ratio hypothesis [MRH]" was formulated that explains the influence of some species on ecosystem processes or services.According to Grime [1998] the influence of dominant plants is proportional to their abundance in a given community.The MRH hypothesis was also confirmed during the experimental studies conducted by Díaz et al. [2007] or Mokany et al. [2008] who stated that the functioning of ecosystem is determined by the morphological, physiological or phonological traits of species that significantly contribute to the production of a biomass.
Grass species play an important role in colonization of such harsh areas [Kompała, Woźniak 2001;Rostański and Woźniak, 2007;Ryś et al., 2016], because they can adapt to various, often extreme, environmental conditions that occur on a post-industrial sites.They frequently possess wide ecological amplitude [Kozłowski et al. 2000].The ability to reproduce vegetatively and to lateral spread via rhizomes as well as dis-persal by wind, enables them to spread into new sites and occupy them [Frey 2000;Rostański, Woźniak, 2007].They frequently play a pioneer role, influencing the soil formation and its physical and chemical properties.Since most of them have a well-developed root system, they also protect the soil against water and wind erosion.Therefore, they are very useful in the reclamation practice.[Frey, 2000;Woźniak, 2010] The aim of the study is [i] to present the differentiation of spontaneous vegetation dominated by chosen grass species occurring on the coalmine spoil heaps; [ii] to identify the main factors responsible for this diversity in the aspect of postindustrial land reclamation; [iii] to find the relations between cover of a dominants and chosen diversity indices and plant biomass.

Site description
The study area is situated in the central part of the Silesian Upland in the mesoregion of the Katowice Upland [Kondracki, 2002].

METHODS
In the experiment, 70 sample plots were established in patches dominated by the chosen grass species.They have a shape of a circle with a radius of 3 m.In each sample plot, the species composition was noted together with the abundance of species evaluated on a 10-degree scale (<1%, 1-5%, 5-10%, and then every 10%).Moreover, smaller plots with a side of 0.5 m were established on 70 sample plots, from which the surface parts of plants (biomass) were collected: separately for a dominant and the rest of species (accompanying species).Shortly after collecting, the fresh mass of species was immediately weighed using a field weight.The samples were then dried at 105ºC in an incubator for 24 hours in the air and weighted again (weight of dry biomass was obtained) as a result of which the weight of dry biomass was obtained (dried at 105°C).
On the basis of the classification results, a synoptic table was prepared.The diagnostic species were determined using as a measure of fidelity phi coefficient of association.The species were classified into ecological groups on the basis of Matuszkiewicz [2001] and Oberdorfer et al. [1990].Species richness, Shannon-Wiener diversity index (H'), Simpson dominance index (D), Evenness index (E) were calculated in JUICE 7. 0. 64 [Tichý, 2002] in order to show the relation between the abundance of a dominants and other characteristics of a community [Tichý, Holt, 2006].In order to present the habitat preferences of the species that constitute the plant commu-nities, the average Ellenberg's indicators values (L, F, R, N) were calculated for each sample plot [Ellenberg et al. 1992] in JUICE 7 0. 64 [Tichý, 2002].The groups of communities were compared using the non-parametric Kruskal-Wallis test [StatSoft, INC. 2011].The names of species were adopted after Mirek et al. [2002].

RESULTS
The hierarchical classification of 70 sample plots enabled us to distinguish 5 groups (clusters) of communities dominated by grasses such as wood small-reed (Calamagrostis epigejos); flattened meadow-grass (Poa compressa); red fescue (Festuca rubra); tall fescue (Festuca arundinacea) and common reed (Phragmites australis) (Table 1) They contain from 8 to 17 plots.The number of species recorded in individual clusters ranges from 23 to 55 species.In individual communities, from 3 to 15 diagnostic species were recorded.High fidelity (phi> 50) to the given community characterized such grass species as: Festuca rubra, Festuca arundinacea, Poa compressa or Phragmites australis.

Taxonomic diversity
The most diverse communities in reference to floristic composition were characterized by high abundance of such species as: flattened meadow grass Poa compressa (H' -1.89), wood small reed Calamagrostis epigejos (H' -1.42), tall fescue Festuca arundinacea (H' -1.45), whereas the communities with dominance of common reed Phragmites australis (H' -0.91) were less diverse.Phragmites australis has the highest average abundance (70%) in patches (Fig. 2a).During studies, the patches with the lower abundance of 20% were rarely recorded.In comparison with other plant communities, the recorded values of Simpson's dominance index were the highest (Fig. 2d).In the patches with such dominants as Calamagrostis epigejos (1), Festuca rubra (3), Festuca arundinacea (4), the mean cover of dominants plant ranges from 30 to 60%.

Dominant species and their impact on species richness, species diversity, evenness, biomass
No statistically significant correlation was found between the abundance of a dominant and the number of species (Table 2).The largest average number of species was recorded in the patches (Fig. 3) dominated by Poa compressa [average 13 species], followed by Festuca arundinacea (average 10 species) and Calamagrostis epigejos (average 10 species).In the remaining panels with the participation of both Festuca rubra and Phragmites australis as dominant species, the total number of species was slightly lower (mean of 6-8 species).
Statistically significant negative correlations were found between the cover of the majority of dominants and the values of Shannon-Wiener diversity index (H') and evenness (E).Only in the case of Poa compressa, there was no statistically significant correlation between the flattened meadow grass cover and the species diversity of the patches.The dominant cover, on the other hand, had a negative effect on the values of the evenness index.The fresh biomass of dominant species was different in the patches of the most frequent vegetation types.The highest fresh biomass was found in the patches dominated by Phragmites australis and Calamagrostis epigejos, whereas the lowest in the patches dominated by Poa compressa (Fig. 4a).There were no statistically significant correlations between the fresh biomass of accompanying species and the dominant cover (Fig. 4b).

Habitat preferences of species occurring in grass-dominated communities on the coalmine spoil heaps
While comparing the grass-dominated communities in terms of habitat preferences of species, it was shown that the communities of Phragmites australis are significantly different due to the habitat preferences of species in terms of humidity, higher nutrient content and attachment of species to shady places.The statistically significant differences were observed between the Phragmites australis community and the Poa compressa and Festuca rubra communities in relation to light (L) and moisture (F) (Fig. 5a, 5b).
The patches with Phragmites australis consist of the species that prefer wet soils and shady places.In contrast, in the floristic composition of patches with Poa compressa and Festuca rubra, the species that are confined to dry soils and open places were found.The statistically significant differences were found in the floristic compo-

DISCUSSION
The data on the biology and habitat preferences of species that spontaneously colonize spoil heaps and form plant communities can be used in the planning reclamation works on various post-industrial wastelands.These species are better adapted to harsh habitat conditions.The use of floristic criteria allowed to distinguish 5 plant types of communities.In the plant communities that develop in post-mining waste dumps, very often one of the most abundant species gives them specific physiognomy and has influence on the diversity of communities.In the case of our   ranging from sunny (Festuca rubra, F. arundinacea) to shaded (Phragmites australis), from dry (Poa compressa, Calamagrostis epigejos) to wet (Phragmites australis), from nitrogen-poor (Poa compressa, Festuca rubra, F. arundinacea) to rich (Phragmites australis), from acidic to alkaline (Poa compressa) substrates.Our results confirmed the results obtained on similar post-industrial sites by Rostański, [2000;2006].Rostański [2000] studying the flora of post-industrial areas, found the considerable diversity of flora in terms of the habitat conditions [moisture and soil fertility, the reaction of the substrate], and the species recorded are often characterized by a wide range of most ecological factors.However, in the flora of postindustrial areas [hard coal mining heaps, sodium heaps], a clear share of mesophilous and moisturedemanding species [Cohn et al. 2001] and of aquatic and rush habitat species [Rostański, 2006] was often observed.According to Rostański [2006] the share of these ecological group of species can be linked to the heterogeneity of the landfilled waste, which is related to the quantity of mineral colloids in the substratum that occur on heaps.

CONCLUSIONS
1. On the basis of the floristic criterion, 5 plant communities dominated by different grass species were distinguished.One species often played a dominant role in them, and cover 20-100% of a patch.2. The communities differ in reference to the species preferences to light, moisture, soil fertility and reaction, which is reflected in the wide variety of microhabitats in the area.

It was shown that the increase in the abundance
of certain grass species, e.g.Calamagrostis epigejos, Festuca rubra, Festuca arundinacea, Phragmites australis, has a significant negative impact on the species richness, species diversity and the uniformity of distribution of species in the plant communities.4. Preliminary analyses revealed that on postmining waste, the biomass production of the dominant species is negatively correlated with biodiversity.5.The knowledge about biology and ecology of grass species, as well as on the assembly rules may be used in the reclamation of degraded areas.

Table 1 .
The characteristics of plant communities occurring on coal-mine spoil heaps of the Katowice Upland Explanations: Ce -community Calamagrostis epigejos, Pc -community Poa compressa, Fr -community Festuca rubra, Fa -community Festuca arundinacea, Pha -community Phragmites australis